CN107810127B

CN107810127B - Cognitive driver assistance with variable alerts for automated vehicles

Info

Publication number: CN107810127B
Application number: CN201680036887.7A
Authority: CN
Inventors: M·H·劳尔; S·兰贝尔蒙特; J·P·阿布斯迈尔
Original assignee: Delphi Technologies Inc
Current assignee: Delphi Technologies Inc
Priority date: 2015-06-24
Filing date: 2016-05-18
Publication date: 2020-11-03
Anticipated expiration: 2036-05-18
Also published as: US9493118B1; EP3313706A1; CN112158132A; EP3313706A4; WO2016209423A1; CN107810127A; EP3313706B1

Abstract

A cognitive driver assistance system (10) includes an object detection device (18), an operator detection device (22), and a controller (20). The object-detection device (18) is operable to detect when an object (16) is proximate to the host-vehicle (14). An operator-detection device (22) is operable to determine when an operator (12) of the host-vehicle (14) is aware of the object (16). The controller (20) is configured to output an alert signal (24) for an operator (12) of the host vehicle (14) when the object detection device (18) detects the object (16). The alarm signal (24) is characterized by a variable alarm intensity (28). The controller (20) is configured to increase the alert intensity (28) when the operator (12) is unaware of the object (16).

Description

Cognitive driver assistance with variable alerts for automated vehicles

Technical Field

The present disclosure relates generally to cognitive driver assistance systems and, more particularly, to systems that output alerts at increased intensity when an operator of a vehicle is unaware of an object approaching the vehicle.

Background

It is known to equip a vehicle with a system that warns an operator of an object approaching the host vehicle driven by the operator, or that takes over control of the host vehicle from the operator to avoid collision with an object approaching the host vehicle. However, in some cases, if the operator is alert and fully aware of the presence/location of an object, an alarm or take over control may be unnecessary and/or tedious.

Disclosure of Invention

Described herein is a cognitive driver assistance system that changes the warning intensity of a warning or changes the degree to which the system takes over control of a host vehicle driven by an operator. The intensity of the alarm and the authority to take over are determined from a measure of the extent to which the operator is aware of a particular object in proximity to the host vehicle or in the path of travel of the host vehicle. For example, if the operator appears tired or inattentive, the alert intensity of the alert may be increased to wake the operator, and/or the amount of operator authority to allow the operator to manually operate the host-vehicle may be limited, such that automated control provided for automated vehicle operation may override the manual control operated by the operator and temporarily take over control of the host-vehicle. Note that the automated operation of the host vehicle includes a wide range of options, from partial automation in which, for example, only speed control is automated to full automation (i.e., autonomous operation) in which the operator of the host vehicle participates in the operation of the host vehicle as much as the passenger.

According to one embodiment, a cognitive driver assistance system is provided. The system includes an object detection device, an operator detection device, and a controller. The object-detection device is operable to detect when an object is proximate to the host-vehicle. The operator-detection device is operable to determine when an operator of the host-vehicle is aware of the object. The controller is configured to output an alarm signal for an operator of the host vehicle when the object detection device detects an object. The alarm signal is characterized by a variable alarm intensity. The controller is configured to increase the alarm intensity when the operator is unaware of the object.

Other features and advantages will appear more clearly on reading the following detailed description of preferred embodiments, which are given by way of non-limiting example only and with reference to the accompanying drawings.

Drawings

The invention will now be described, by way of example, with reference to the accompanying drawings, in which:

FIG. 1 is a diagram of a cognitive driver assistance system according to one embodiment;

FIG. 2 is an exemplary traffic situation experienced by the system of FIG. 1, according to one embodiment;

FIG. 3 is an exemplary traffic situation experienced by the system of FIG. 1, according to one embodiment;

FIG. 4 is an exemplary traffic situation experienced by the system of FIG. 1, according to one embodiment;

FIG. 5 is an exemplary traffic situation experienced by the system of FIG. 1, according to one embodiment; and

FIG. 6 is an exemplary traffic situation experienced by the system of FIG. 1, according to one embodiment.

Detailed Description

Fig. 1 shows a non-limiting example of a cognitive driver assistance system 10 (hereinafter system 10). In general, the system 10 is configured to assist the operator 12 in manually operating the host-vehicle 14 based on the cognitive state of the operator 12 with respect to the object 16 proximate to the host-vehicle 14. As described in more detail below, the level of assistance provided by the system 10 to the operator 12 may vary from as simple as something that illuminates an alarm indicator when an object 16, such as another vehicle, is present in the so-called blind spot of the operator 12, to something that is complex to take over full control of the host-vehicle 14 (i.e., completely override the operator 12) to avoid a collision with the object 16. The system 10 is described herein as having an advantage over prior art driver assistance systems in that the system 10 takes into account the cognitive state of the operator 12 so that the level of assistance necessary to assist the operator 12 can be determined. For example, if the operator 12 is not aware of a particular object associated with an alarm, the loudness of the audible alarm indicator may be increased.

Accordingly, the system 10 includes an object-detection device 18 operable or operable to detect when an object 16 is proximate to the host-vehicle 14. The object-detection device 18 may include, but is not limited to, a camera, a radar unit, and/or a lidar unit suitably configured to detect various objects around the host-vehicle 14. By way of further example and not limitation, a camera may be used to capture images of objects such as other vehicles, buildings, road signs, pedestrians, lane markings, road hazards, architectural area markings, and other objects proximate to the host-vehicle 14 or disposed about the host-vehicle 14.

The information from the camera may be combined with detection information from the radar unit and/or lidar unit to form a two-dimensional or three-dimensional map of various objects in the area surrounding the host-vehicle 14. Information collected by operation of the object detection device 18 may be provided to the controller 20 for further analysis. That is, the object-detecting device 18 may not be able to actually determine the location or classification of the object 16 by itself. Typically, this task falls to the controller 20, but this is not a requirement of the system 10. In other words, either the object detection device 18 or the controller 20 may perform the data or signal processing necessary to determine the relative position or classification of the object 16.

The object detection device 18 may also include a vehicle-to-vehicle (V2V) transceiver. If the object 16 is another vehicle that is also equipped with a V2V transceiver, the system 10 can notify the operator 12 that another vehicle is approaching, even if the operator 12 cannot see the other vehicle because the line of sight between the two vehicles is obstructed.

System 10 may be configured to classify object 16 based on signals received from object detection device 18. For example, if the object 16 is a pedestrian walking perpendicular to the direction of travel of the host-vehicle 14, the signal from the lidar unit or the radar unit may have a unique signature due to the alternating motion of the legs of the pedestrian. Similarly, radar returns reflected from a moving vehicle may include minor fluctuations due to rotational motion of the wheels of the moving vehicle. This information may optionally be combined with other known image processing techniques to classify the object as, for example, a pedestrian, another vehicle, a building area marker, an animal, a road obstacle, etc. It is also possible to use information about the movement of the object 16 to classify the object 16 and to assign parameters to the object 16 related to the movement, such as speed, trajectory, acceleration/deceleration, etc. As such, the object-detection device 18 is operable to detect when the object 16 is approaching, i.e., approaching, or otherwise approaching the travel path of the host-vehicle 14 within the travel path.

It should be apparent to those skilled in the art that the controller 20 may include a processor (not specifically shown), such as a microprocessor or other control circuitry, such as analog and/or digital control circuitry, including an Application Specific Integrated Circuit (ASIC) for processing data. The controller 20 may include memory (not specifically shown), including non-volatile memory, such as electrically erasable programmable read-only memory (EEPROM), for storing one or more routines, thresholds, and collected data. The one or more routines may be executed by the processor to perform steps for processing signals received by the controller 20 to operate the host-vehicle 14 as described herein.

The system 10 also includes an operator-detection device 22 operable or useful for determining when the operator 12 of the host-vehicle 14 is aware of the object 16. Operator-detection device 22 may include a camera configured to take images of operator 12 in the visible and/or infrared spectrum, and a light source that emits light in appropriate portions of the spectrum to enable the camera to see operator 12 regardless of ambient lighting conditions. The image of the operator 12 may be processed/analyzed by the operator-detection device 22 or the controller 20 to, for example, determine whether the eyes of the operator 12 are closed (closure indicating that the operator 12 is sleeping), determine whether the head pose of the operator is changing in a manner that indicates fatigue of the operator 12, and/or determine whether the eye gaze direction of the operator 12 indicates that the operator 12 has or has not seen the object 16. Image processing techniques are well known that determine the cognitive state of the operator 12, such as where the operator is looking (eye gaze direction) and/or whether the operator 12 is alert (eye blinking, head pose change).

If it is observed or determined via operator detection device 22 that the eye gaze direction of operator 12 is not in a direction that allows operator 12 to see object 16 directly, via mirrors, or via peripheral vision, then that may be an indication that operator 12 is not aware of object 16. If the system 10 determines that the operator 12 is unaware of the object 16 and that the relative position of the object 16 with respect to the host-vehicle 14 is such that a collision or conflict is likely, the system 10 may take various actions to assist the operator 12 in avoiding a collision or conflict with the object 16. As used herein, a collision with an object 16 means that the trajectories of the host-vehicle 14 and the object 16 are such that a collision may occur if either is not altered. By way of example and not limitation, a conflict occurs when the object 16 is another vehicle traveling in a lane beside the host-vehicle 14 and in the blind spot of the operator 12, and the operator 12 operates the host-vehicle 14 in a manner that causes the other vehicle to turn to avoid a collision. That is, no collision occurs, but a collision is avoided due to the action of another vehicle to avoid a collision with the own vehicle 14.

In one embodiment of the system 10, the controller 20 may be configured to output an alert signal 24 to the operator 12 of the host-vehicle 14 when the object 16 is detected by the object-detection device 18. The alert signal 24 is provided to an alert device 26. the alert device 26 may include, for example, a visual alert device (e.g., an illuminated indicator or highlight region of a reconfigurable display), an audible alert device (e.g., a buzzer or speaker), and/or a tactile alert device (e.g., a vibrating device located in the seat or steering wheel of the host vehicle 14).

In the present embodiment, the alarm signal 24 is characterized by a variable alarm intensity 28. For example, the intensity of the alert of the alerting device 26 can be increased by increasing the loudness of the audible alert, increasing or blinking the highlighting intensity of the visual alert, and/or combining two or more of the types of alerts available from the alerting device 26 simultaneously. It is envisaged that the alarm intensity 28 may be initially set to zero or none, such that no alarm is given, i.e. no alarm means are activated, and increasing the alarm intensity 28 corresponds to changing from no alarm being given to some alarm being given. That is, not requiring an increase in the alert intensity 28 necessarily means that the initial alert intensity is not zero or none.

The system 10 described herein is an improvement over prior art systems in that the controller 20 is configured to increase the alert intensity 28 when the operator 12 is not aware of the object 16. That is, rather than issuing an alarm at the same intensity when the operator 12 is alert and aware of the object 16, and when the operator 12 is drowsy and not aware of the object 16, the system 10 varies the alarm intensity 28 based on the cognitive state of awareness of the operator 12, and in particular whether the operator 12 is aware of the presence of the object 16. Several specific traffic scenarios are described later in this document, which will make the advantages of the system 10 described herein readily apparent.

In another embodiment, the system 10 includes a control override device 30 operable to limit operator authority 32 of the operator 12 while the operator 12 is driving the host-vehicle 14. As used herein, "limiting operator authority" means limiting the ability of an operator to manually control some aspect of the operation of the host-vehicle 14, i.e., preventing or to some extent limiting the ability of the operator 12 to fully manually control the host-vehicle 14. The controller 20 instructs to limit the operator authority 32 by setting the operator authority 32 to an authority other than "full authority". The controller 20 replaces manual control of the host vehicle 14, in whole or in part, by activating the control override device 30. By way of example and not limitation, the control override device 30 may include one or more of a steering override device 34, an accelerator override device 36, and a brake override device 38.

By way of further example, the steering override device 34 may be operated by the controller 20 to override steering wheel inputs from the operator 12 to prevent the operator 12 from making a lane change because the object detection device 18 has detected another vehicle in the vicinity of the host-vehicle 14 and would cause a collision if a lane change were made. Similarly, the steering override device 34 may be operated by the controller 20 to steer the host vehicle 14 in place of the operator 12 holding the steering wheel in a fixed position so as to avoid certain debris in the driving lane of the host vehicle 14.

In one non-limiting configuration, the steering override device 34 may apply a force to the steering system of the host-vehicle 14 that physically overrides any attempt by the operator 12 to affect steering of the host-vehicle 14. This configuration may be preferred when there is an uninterruptible mechanical connection between the steering wheel used by the operator 12 to steer the host-vehicle 14 and the steered wheels (e.g., front wheels) of the host-vehicle 14. In alternative constructions, the steering system may include a steering clutch operable to mechanically decouple the steering wheel from the front wheels, or the steering system may be a steer-by-wire type system, such that steering wheel rotations by the operator 12 may be readily ignored.

As described in the examples above and in other examples given below, where the operator 12 is unaware of the presence of the object 16 and a collision or conflict may occur, what the operator 12 indicates for the desired operation of the host-vehicle 14 may be limited or ignored in favor of operating the host-vehicle 14 in accordance with the controller 20. That is, the controller 20 may be configured to operatively control the override device 30 to override the operator 12 in accordance with the operator authority 32 to avoid conflict with the object 16 when the operator 12 is unaware of the object 16.

Fig. 1 shows that there are two levels of permissions: full rights and restricted rights. However, many levels of partial authority are contemplated, and the specified level will be based on periodic observations of the operator by the operator detection device 22. It is also contemplated that the actions of providing an alert to operator 12 and overriding the effort of operator 12 to manually control the host-vehicle may be combined. For example, when the controller 20 determines that the operator 12 is not aware of the object 16, the controller 20 may first activate the alert device 26 at a lower level of alert intensity 28, and then increase the alert intensity 28 to a higher level before the controller 20 determines to limit the operator's privileges 32 by activating the control override device 30.

The controller 20 may also be configured to determine an operator behavior characteristic 40 of the operator 12 based on previous traffic maneuvers 42 performed by the operator 12. As used herein, the operator behavior characteristics 40 are used to characterize the driving habits or driving style of the operator 12. The operator behavior characteristics 40 may include or relate to, but are not limited to, how fast the operator 12 is driving relative to the published speed limit, i.e., how many miles per hour the operator 12 typically travels over or under the published speed limit. Other examples include how fast the operator 12 accelerates from a stop, how fast the operator 12 decelerates to a stop when approaching an intersection, how often the operator 12 looks at the rear and side mirrors or otherwise scans the area near the host-vehicle 14, and how long the operator 12 activates the turn signal before making a lane change. Controller 20 may calculate a previous object detection score 44 based on a percentage of time that operator 12 appears to have actually detected the presence of an object detected and/or tracked by system 10.

The controller 20 may also be configured to determine a collision risk 46 indicative of a probability of collision between the host-vehicle 14 and the object 16. As used herein, "probability of collision" includes, but is not limited to, an estimate of the probability of an actual collision, and/or an estimate of the probability of some change in the observed behavior of the operator 12, the host-vehicle 14, or some other vehicle (or object) in proximity to the host-vehicle 14 necessary to avoid a collision. The controller 20 may determine the alert intensity 28 and/or the operator authority 32 based on the previous traffic maneuver 42 and the risk of collision 46. Examples of various traffic situations are provided next to further explain how the controller 20 described thus far operates to assist the operator 12 in operating the host-vehicle 14.

FIG. 2 illustrates a non-limiting example of a truck 48 traveling on an approaching road 50 where the truck 48 is traveling slower than the host-vehicle 14 and in the same lane, and thus the host-vehicle 14 may be about to overtake the truck 48. In this example, the object 16 is another vehicle 52 that is traveling faster than the host-vehicle 14 and is in a lane in which the host-vehicle 14 may be used beyond the truck 48. The operator-behavior characteristics 40 observed by the controller 20 may include an indication that the operator 12 tends to slow down as it approaches another vehicle in front of the host-vehicle 14, rather than maintaining a constant speed and making a lane change in time. Also, while the operator 12 is performing a lane change, the controller 20 may be configured to measure the time it takes the operator 12 to complete the lane change. That is, the operator behavior characteristics 40 may include a lane change duration (not shown) stored in the controller 20 and based on a previous lane change (previous traffic maneuver 42) performed by the operator 12.

Thus, the risk of collision 46 may include an estimate of the time to collision 54 left for a collision between another vehicle 52 and the host-vehicle 14. The controller 20 may be configured to determine the alert intensity 28 based on a comparison of the lane change duration to the time left 54 for the conflict to occur. For example, the loudness of the alert may increase as the time left 54 (FIG. 1) for the conflict to occur decreases while the operator 12 appears to be considering lane changes. Alternatively, or in combination with changing the alert intensity 28, the controller 20 may limit the operator authority 32 when another vehicle 52 is approaching too quickly and prevent the operator 12 from changing lanes beyond the truck 48 to avoid significant collisions.

With continued reference to another example of FIG. 2, the object 16 may be another vehicle 56 traveling in a so-called blind spot of the operator 12 of the host vehicle. The operator-behavior characteristics 40 may include a measure of a lane-position-change 58 of the host-vehicle 14 as the host-vehicle 14 is steered by the operator 12. The controller 20 may also determine similar characteristics of the other vehicle 56 based on a previous travel path 59 of the other vehicle 56. The determination of the risk of collision 46 by the controller 20 may include determining a probability of collision (not specifically shown) between another vehicle 56 and the host-vehicle 14 when the other vehicle 56 is alongside the host-vehicle 14. That is, the controller 20 may estimate the risk of a collision occurring due to the host-vehicle 14 and/or another vehicle 56 traveling.

If the vehicle is not traveling and the operator 12 is aware of another vehicle 56, the controller 20 may issue an alert with a lower intensity blind spot alert. However, if the estimated probability of collision is greater than a risk threshold (not shown) stored in the controller 20, the controller 20 may increase the alert intensity 28 by making the audible alert louder and/or displaying a message on the display to indicate that the operator 12 is over-traversing when another vehicle 56 approaches the host-vehicle 14. Alternatively, or in combination with increasing the alert intensity 28, the controller 20 may limit the operator authority 32 and prevent the operator 12 from traversing by providing some steering damping to the steering system via the steering override device 34 in an attempt to reduce the value of the lane position change 58.

FIG. 3 illustrates another non-limiting example in which the host-vehicle 14 is traveling on a road 60 and the object 16 is stationary at a location in the path of travel 62 of the host-vehicle 14. For example, the object 16 may be debris 64 dropped from a truck (not shown) previously traveling on the roadway 60. The controller 20 may be further configured to estimate a time remaining for the host-vehicle 14 to impact the object 16 (debris 64) (FIG. 1). It is contemplated that the time to impact 66 is calculated based on an assumption that the host vehicle continues along the travel path 62 without a change in speed or trajectory.

If the time left until impact occurs is long, such as greater than five seconds, the controller 20 may set the alert intensity 28 to be relatively low and illuminate only indicators that may be observed by the operator 12 of the host-vehicle 14. However, if the time left for impact 66 is less than a time threshold (not shown) stored in the controller 20, such as less than 2 seconds, and/or the gaze direction of the operator 12 indicates that the operator is not aware of the debris 64, the controller 20 may increase the alert intensity 28 by additionally activating an audible alert and/or activating a haptic device to vibrate the seat of the operator 12. Alternatively, or in combination with increasing the alert intensity 28, the controller 20 may restrict the operator's authority 32, activate the steering override device 34, and operate the steering override device 34 to avoid the object 16 when the operator 12 is unaware of the object 16.

With continued reference to fig. 3, the object 16 may be a pedestrian or animal 68 proximate the travel path 62 of the host-vehicle 14. The controller 20 may be further configured to estimate a time-to-impact 66 based on the speed or relative trajectory of the pedestrian or animal 68 and the host-vehicle 14. If the object is a pedestrian (e.g., a person), the gaze direction of the pedestrian may be determined using the camera of the object detection device 18 to determine whether the pedestrian is aware of the host-vehicle 14. The controller 20 may increase the alert intensity 28 when the time remaining for impact 66 is less than a time threshold (e.g., 2 seconds) and the operator 12 is not aware of the pedestrian or animal 68. Alternatively, or in combination with increasing the alert intensity 28, the controller 20 may activate or operate the brake override device 38 and/or the accelerator override device 36 to stop the host vehicle 14 when the operator 12 is unaware of the pedestrian or animal 68.

FIG. 4 illustrates another non-limiting example in which the host-vehicle 14 is traveling on a road 70 and the object 16 is a traffic light 72 on a path of travel 74 of the host-vehicle 14. The controller 20 (fig. 1) may be further configured to: if the operator 12 does not initiate braking and stop the host-vehicle at the intersection 78, the time-to-violation 76 (FIG. 1) of the traffic light 72 by the host-vehicle 14 is estimated, and the alert intensity 28 is increased when the time-to-violation 76 is less than a time threshold (e.g., three seconds) and the operator 12 is not aware of the object 16 (traffic light 72). It is contemplated that the threshold time may vary based on the current travel speed of the host-vehicle 14, e.g., for greater travel speeds, the threshold time may be greater. Alternatively, or in combination with increasing the alert intensity 28, the controller 20 may activate or operate the brake override device 38 and/or the accelerator override device 36 to stop the host-vehicle 14 prior to entering the intersection 78 when the operator 12 is unaware of the traffic light 72.

Fig. 5 shows another non-limiting example in which the host-vehicle 14 is traveling on a road 80 and the object 16 is another vehicle 82 approaching an intersection 84 via an intersection lane 86 of the intersection 84. In this situation, the host vehicle 14 is stopped and ready to enter the intersection 84 to begin turning or straight-driving. There is no need to stop traffic on the intersecting lane 86, so if the host-vehicle 14 continues to enter the intersection 84 before another vehicle 82 reaches the intersection 84, the time left until a collision occurs (FIG. 1) and/or the time left until a collision 66 should be considered. Preferably, the control override device 30 includes an accelerator override device 36, and the controller 20 may operate the accelerator override device 36 to prevent the operator 12 from entering the intersection 84 when the operator 12 is unaware of another vehicle 82. Alternatively, or in combination with operating the brake override device 38, the controller 20 may increase the alert intensity 28 when the operator 12 is unaware of the other vehicle 82, and/or the controller 20 may activate or operate the accelerator override device 36 to move the host-vehicle 14 away from the illustrated position.

An alternative scenario associated with fig. 5 is when the host vehicle 14 is stopped and ready to turn right as shown, and there is a bicycle (not shown) approaching the intersection behind the host vehicle 14. The system 10 may determine that the bicycle would collide or conflict with the host-vehicle 14 if a right turn were initiated and increase the alert intensity 28 before the conflict occurs.

Fig. 6 shows another non-limiting example in which the host vehicle 14 is traveling on a road 90 and the object 16 is a train 92 approaching a railroad highway intersection 94 of the road 90. The control override device 30 advantageously includes a brake override device 38, and the controller 20 operates the brake override device 38 to prevent the operator 12 from stopping at the railroad road intersection 94 when the operator 12 is unaware of the train 92. It is contemplated that the host vehicle 14 may need to continue to travel and contact the stopped vehicle 96 to avoid being hit by the train 92. Alternatively, or in combination with operating the accelerator override device 36, the controller may increase the alert intensity 28 when the operator 12 is not aware of the train 92, and/or the controller 20 may activate or operate the brake override device 38 to maintain the host vehicle 14 at the illustrated position.

Accordingly, a cognitive driver assistance system (system 10) and a controller 20 for the system 10 are provided. The controller 20 is capable of determining the cognitive state of the operator 12, such as determining when the operator is alert and aware of objects disposed around the host-vehicle 14, and when the operator is not alert and aware of a particular object (object 16). If the controller 20 determines that some action is being taken by the operator 12 other than the action that the operator is doing when the operator is unaware of the object 16, the controller may increase the alert intensity 28 of the alert provided to the operator 12 and/or take over control of the host-vehicle 14 in whole or in part and operate the host-vehicle 14 to avoid a collision or conflict with the object 16.

While the present invention has been described with respect to the preferred embodiments thereof, it is not intended to be limited thereto, but rather only to the extent set forth in the following claims.

Claims

1. A cognitive driver assistance system (10), the system (10) comprising:

an object detection device (18) operable to detect when an object (16) is proximate to the host-vehicle (14);

an operator-detection device (22) operable to determine when an operator (12) of the host-vehicle (14) is aware of the object (16); and

a controller (20) configured to output an alert signal (24) for the operator (12) of the host-vehicle (14) when the object (16) is detected by the object-detection device (18), the alert signal (24) characterized by a variable alert intensity (28), wherein the controller (20) is configured to increase the alert intensity (28) when the operator (12) is unaware of the object (16);

wherein the controller (20) is further configured to: determining an operator behavior characteristic (40) of the operator (12) based on a previous traffic maneuver (42) performed by the operator (12); and determining the alert intensity (28) based on the previous traffic maneuver (42) and a collision risk (46) indicative of a probability of collision between the host-vehicle (14) and the object (16); and wherein the object (16) comprises another vehicle (52), the operator-behavior characteristic (40) comprises a lane-change duration based on a previous lane change performed by the operator (12), the collision risk (46) comprises an estimate of a time-to-collision (54) remaining for a collision between the other vehicle (52) and the host-vehicle (14), and the alert intensity (28) is determined in part by comparing the lane-change duration and the time-to-collision (54).

2. The system (10) in accordance with claim 1, wherein the operator-behavior characteristic (40) includes a lane-position-change (58) of the host-vehicle (14) as the host-vehicle (14) is steered by the operator (12), the risk of collision (46) includes a probability of collision between the host-vehicle (14) and the other-vehicle (52) when the other-vehicle (52) is beside the host-vehicle (14), and the alert intensity (28) is determined in part by comparing the probability of collision to a risk threshold.

3. The system (10) in accordance with claim 1, wherein the object (16) comprises a stationary object at a location in a travel path (62) of the host-vehicle (14), said controller (20) further configured to estimate a time-to-collision (66) of the host-vehicle (14) with the object (16), and increase the alert intensity (28) when the time-to-collision (66) is less than a time threshold and the operator (12) is unaware of the object (16).

4. The system (10) in accordance with claim 1, wherein the object (16) includes a pedestrian or animal (68) proximate a travel path (62) of the host vehicle (14), the controller (20) further configured to estimate a time-to-impact (66) of the host vehicle (14) with the pedestrian or animal (68), and increase the alert intensity (28) when the time-to-impact (66) is less than a time threshold and the operator (12) is not aware of the pedestrian or animal (68).

5. The system (10) in accordance with claim 1, wherein the object (16) comprises a traffic light (72) of a travel path (62) of the host-vehicle (14), the controller (20) further configured to estimate a time-to-violation (76) of the traffic light (72) by the host-vehicle (14), and increase the alert intensity (28) when the time-to-violation (76) is less than a time threshold and the operator (12) is unaware of the object (16).